Five Molecules We Would Take to a Remote Island

(1)

Five Molecules We Would Take to a Remote Island

Thomas U. Mayer1,* and Andreas Marx¹

lKonstanz Research School Chemical Biology (KoRS-CB), University of Konstanz, Universitatsstr. 10, 78457 Konstanz, Germany 'Correspondence: Thomas.u.mayer@uni-konstanz.de

Which five molecules would you take to a remote island?

If

you imagine yourself as a castaway on an island you might pick water, glucose, penicillin, and ethanol in combination with aspirin. However, as a scientist, you may ask yourself which molecules impressed you most by their chemical or biological property, their impact on science, or the ingenuity and/or serendipity behind their discovery. Here, we present our personal short list comprising FK506, colchicine, imatinib, Quimi-Hib, and cidofovir. Obviously, our selection is highly subjective and, therefore, we apologize up front to our colleagues for not mentioning their favorite compounds.

FK506

FK506 is a fascinating molecule because of both its medical impact and its intriguing mode of action. FK506 is a macrolide with immunosuppressant activity produced by the bacterium Streptomyces tsukubaensis, giving FK506 the alternative name Tacrolimus ([sukuba macrolide immyno§uppressant). FK506 was identified in 1984 in the course of a systematic search for novel immunosuppressant agents by a Japanese team headed by T. Goto, T. Kino, and H. Hatanaka (Goto et aI., 1987; Tanaka et aI., 1987). Ten years later, FK506 was already approved for use in liver transplantation, and by now is a widely applied suppressor of allograft rejections. Its medical impor- tance fueled the search for the cellular target of FK506 and succeeded in the identification of FKBP (FK506-binding protein), a peptidyl-prolyl isomerase (PPlase) (Harding et aI., 1989; Siekierka et aI., 1989, 1990; Standaert et aI., 1990).

PPlases increase the rate of protein folding by catalyzing the cis-trans isomer- isation of peptide bonds N-terminal to proline residues. FK506 inhibits FKBP by a substrate mimicking mechanism, whereby the a-keto amide of FK506 serves as a surrogate for the twisted amide of a bound peptide substrate (Rosen et aI., 1990). However, the loss of PPlase activity is not the basis of the T cell inhibition by FK506. Instead, the complex of FK506 and FKBP inhibits the phosphatase calcineurin (CaN), also known as PP2B, thereby preventing the dephosphorylation of NF-AT (nuclear factor of activated T cells) that is required for the expression of early T cell activation 556

genes, e.g. IL-2 (Liu et aI., 1991). Neither FK506 nor its target FKBP by itself can bind to CaN. Thus, rather than by simply inhibiting the function of its target protein, FK506 induces

a

gain-of-function by imparting new binding properties to FKBP, which results in the termination of the T cell receptor-initiated immune response at the level of CaN (Figure 1).

For those not yet convinced by FK506's claim to be taken to a remote island, it might be noteworthy that the story of immunosuppressants is even more complicated because the small molecules cyclosporine A (CsA) and rapamycin suppress T cell activation by a related, yet distinct pathway (for review, see Powell and Zheng [2006]). CsA, a cyclic undeca- peptide produced by the fungus Beauveria nivea, binds to cyclophilin (CpN), a PPlase unrelated in sequence to FKBP. However, both the complexes FK506/FKBP and CsAlCpN converge on calcineurin and thus suppress T cell activation by a common mechanism, Le., inhibiting the dephosphorylation of NF-AT by CaN. On the contrary, rapamycin, a macrolide chemically related to FK506 and produced by the bacterium Streptomyces sirolimus, binds to FKBP-Iike FK506, but the FKBPI rapamycin complex interacts with mammalian target of rapamycin (mTOR), a serine-threonine protein kinase, instead of CaN and exerts its immunosuppressive activity by inhibiting the response to IL-2 rather than its expression. Thus, FK506 and rapamycin share a common structural element mediating binding to FKBP while having different effector elements, resulting in the inhibition of different signaling pathways (Bierer et aI., 1990).

FK506's mode of inhibition is not only intriguing and holds several unexpected twists, but it is also an excellent example of how the combined effort of chemists, biologists, and physicians yielded a highly potent small molecule that is not only invaluable for basic research but also indispensable for modern medicine.

Colchicine

The story of colchicine started in ancient times, when Jason and his famous Argo- nauts were sent to Colchis, the old kingdom of Georgia, in Asia to retrieve the Golden Fleece, which-according to historians-was nothing but a mass of golden crocus. In Europe, extracts of the golden crocus was in great demand to treat gout, which Hippocrates coined podagra (podos = foot; agra

=

seizure) due to the fact the first inflammations typi- cally attack the foot. Two millennia later, we now know that the active ingredient of the mystic Asian pharmacophore is colchicine, whose value as an anti-inflammatory agent for the treatment of gout is still acknowledged (Figure 2). While the isolation of the alkaloid colchicine from Colchicum autumnale (Autumn crocus also known as "Meadow saffron") was first achieved by Pelletier and Caventou in 1820 (Pelletier and Caventou, 1820), its structure was not determined until 1940 (Cohen et aI., 1940). Notably, the cytological effect of colchicine was already described in 1889 by Biaggio Per- nice, a physician from Palermo (Pernice and Caventou, 1889). He noted "quasi tutti gli elementi in cariocinesi," Le., a dramatic accumulation of mitotic cells in the gastric and intestinal mucosa of Zuerst ersch. in : Chemistry and Biology ; 17 (2010), 6. - S. 556-560

DOI : 10.1016/j.chembiol.2010.06.002

Konstanzer Online-Publikations-System (KOPS) URN: http://nbn-resolving.de/urn:nbn:de:bsz:352-125462

(2)

dogs treated with high doses of colchicine. Subsequent studies revealed that colchicine had major effects on the

.~

.t¥FK506 FKBP

signa!

t

inactive

mitotic spindle (Dustin, 1934;

Eigsti, 1938; Levine, 1945) and it was in fact colchicine's antimitotic effect that was used to unambiguously deter- mine that 46 is the normal human diplOid number of chromosomes rather than the previously believed 48 (fjio and Levan, 1956). Thus, colchicine has a one-of-a- kind historical record as a drug and a biological probe- even before its target was known. The advent of radio- isotopes finally enabled

.XA~

~/;V

^CsA ^{1 ___ -}

,../ ... 1

tively (Panda, et aI., 1995, Rav- elli, et aI., 2004). Colchicine binds to a single site on ~

tubulin (colchicine site), which is shared by a large number of molecules structurally unrelated to colchicine, e.g., podo- phyllotoxin, combrestatin-A4.

The high specificity of colchicine for tubulin fueled the search for an endogenous molecule that regulates microtubule dynamics via the colchicine site. However, identified candidates have not been widely accepted. Thus, the story of colchicine has still to be continued.

.\i, ___ ^~ //

h

^K> i n a c t i v e " , . ,

I +

translation, autophagy -_._\\--....

E ...

ribosome biogenesis -J

mitochondrial metabolism ,.

-t

"

rapa

4-[(4-Methylpiperazin-1- yl)methyl]- N-[4-methyl- 3-[(4-pyridin-3-

ylpyrimidin-2-yl)amino]

phenyl]benzamid

\ . + t ',,--

'~----,/ ... ---~- Edwin Taylor and Gary Borisy

to identify' tubulin as the cellular target of colchicine.

Colchicine-³H, prepared by methylation of colchiceine with diazomethane in tritiated water, was found in a nonco- valent complex with a macro- molecule that showed a corre- lation with the presence of microtubules (Borisy and Tay-

Figure 1. Simplified Diagram of T Cell Receptor Mediated Signal Transduction Pathway Resulting in the Expression of Interleukin 2 Receptor activation triggers a transient rise in intracellular calcium levels, and thus the activation of the phosphatase calcineurin (CaN). Dephosphorylated NF-AT enters the nucleus and induces the expression of 1l2. Note that FK506 and rapamycin share common structural elements mediating binding to FKB, yet have distinct effector elements. FK506 binding protein, FKBP;

cyclophilin, CpN; mammalian target of rapamycin, mTOR.

This is better known as glee- vec, glivec, imatinib, STI571, or the magic bullet for the treatment of chronic myelog- enous leukemia (CML) (Fig-

lor, 1967a, 1967b; Taylor, 1965). Using cOlchicine-³H and GTP-³H, Taylor and colleagues not only identified a 120kDa dimer (Le., the o:-/~-tubulin-dimer) as the target of colchicine, but also discovered that only one of the two GTP molecules bound per dimer is exchanged (Moh ri, 1968; Shelanski and Taylor, 1967; Wei- senberg et aI., 1968). Today, we know that the cycle of GTP hydrolysis and nucleotide exchange by ~-tubulin forms the basis for the key characteristic of microtubules: dynamic instability. This phenom- enon describes the stochastic bidirec- tional switches of microtubules between phases of growth and shrinkage that allows them to quickly explore cellular space, to direct cell movement, and to segregate chromosomes (Desai and Mitchison, 1997).

Does the identification of tubulin finally disclose all mystery of colchicine? Not really. It is still an open question of how colchicine's mode of action explains its major clinical use in gout. Gout is caused by the deposition of uric acid crystals in joints and surrounding tissues. Colchicine might prevent the formation of uric acid

crystals by modulating the pH ofthe tissue, act as an anti-inflammatory agent by sup- pressing the activity of neutrophils, or inhibit the invasion of immune cells into damage tissue-to name a few potential explanations. Unlike other microtubule- binding drugs, colchicine is not used in cancer chemotherapy, probably due to its high toxicity (Jordan and Wilson, 2004).

However, colchicine is widely applied in plant breeding to produce diploid gametes resulting in plants with stronger flower colors. The binding of colchicine to tubulin has several intriguing features. It is a two-step process that includes a fast reaction yielding a low-affinity complex followed by a slow, unimolecular step to form the poorly reversible colchicine- tubulin (TC) complex (Engelborghs, 1998).

The TC complex incorporates at the end of growing microtubules where it prevents curved tubulin from adopting a straight conformation. This conformational con- straint prevents the stabilization of lateral contacts between tubulin subunits resulting in loss of dynamic instability and the de polymerization of microtubules at low and at high TC concentrations, respec-

ure 3). CML is a malignancy of a pluripotent hematopoietic stem cell. Ninety-five percent of CML patients are positive for the Philadelphia (Ph) chromosome, which was named after the location of the two research insti- tutes where it was first discovered and described (Nowell, 1962; Rowley, 1973).

The Ph-chromosome is the result of t(9;22)(q34;q11), i.e., the reciprocal translocation between the large arms of chromosome 9 and 22. The molecular consequence of this translocation is a replacement of the first exon of the proto- oncogene c-Abl with sequences from the BCR ("break point cluster region") (Heist- erkamp et aI., 1985; Konopka et aI., 1984;

Shtivelman et aI., 1985). c-Abl encodes a nonreceptor tyrosine kinase that has tightly controlled activity in normal cells.

In contrast, the protein product of the Bcr-Abl fusion shows constitutive tyrosine kinase activity. Thus, BCR-Abl allows cells to proliferate in a cytokine-independent manner. Furthermore, BCR-Abl's inhibitory effect on DNA repair induces genomic instability, which might be causative for the feared blast crisis in CML.

The story of imatinib began in the early 1990s at Ciba-Geigy (now Novartis)

(3)

A B

c

Figure 2. Colchicine and Its Effect on Mitosis

(A) First description of colchicine's effect on cells in mitosis by the Italian physician Biaggio Pernice, 1889 (image was obtained through the Biodiversity Heritage Library).

(B) Chemical structure of colchicine.

(C) Electron micrograph of sperm tail preparation showing the characteristic 9+2 microtubule structure, Le., two central microtubule doublets surrounded by nine outer doublets (Shelanski and Taylor, 1967).

when a large number of 2-phenylamino- pyrimidine derivatives were screened for inhibitory activity against protein tyrosine kinases (PTKs). The originally identified lead compound inhibited Abl-kinase in vitro; however, its primary cellular target was the platelet-derived growth factor receptor (PDGF-R) PTK. Using this compound as a lead structure, rational drug design led to the identification of imatinib, which inhibited the autophos- phorylation of Abl and PDGF-R with equal potency (Buchdunger et aI., 1996). Impor- tantly, colony formation assays with bone marrow cultures from CML patients demonstrated that imatinib inhibited the growth of BCR-Abl positive cells to an extent of 92%-98% (Druker et aI., 1996).

Normal colonies were minimally affected, despite the fact that imatinib inhibited in vitro BCR-Abl and c-Abl with an identical ICso. Extensive animal testing, including a CM L -mouse model, confirmed imatinib's high efficiency as well as low 558

toxicity and paved the way forthe first clinical trials just few years after its discovery.

Today, imatinib mesylate is approved as first-line treatment for CML. Several patients, however, develop drug resis- tance partly due to amplification or muta- tion of the BCRlABL gene, suggesting that combined therapies including imatinib should be considered (Wailer, 2010).

Imatinib is an ATP-competitive inhibitor, raising the question of how this molecule can be selective for the PTKs Abl, PDGF-R, and c-kit, another PTK. The crystal structure of Abl in complex with an imatinib variant revealed that the inhibitor recognizes the inactive conformation of the activation loop of Abl (Schindler et aI., 2000). For most kinases the activation loop controls catalytic activity. Impor- tantly, the "active" conformation of the loop, stabilized by phosphorylation, is highly similar in all known structures of active kinases, while the "inactive" conformation displays great diversity. Thus,

despite the highly conserved nucleotide- binding pocket of kinases, the characteristic conformation of the activation loop in its inactive state allows imatinib to be highly selective for its targets while being essentially inactive against other PTKs and serine/threonine kinases.

Why do we consider Imatinib worth taking to a remote island? First, it is an excellent example of the synergistic research betweenacademia and industry.

The combined efforts of Brian Druker, Nicholas Lydon, and Charles Sawyers were awarded with the Lasker-DeBakey Award in 2009. Second, its intriguing mode of action identified kinases as drug- gable, despite the high sequence conser- vation of their catalytic center. Third and most importantly, it shifted the role of physicians from checking blood counts and delivering cruel news to efficiently treating CML patients.

Quimi-Hib

Immunogens used in many vaccination strategies derive from constituent parts of the pathogen. For instance, the vaccine against Haemophilus influenzae type b (Hib), a pathogen that causes bacterial meningitis and pneumonia that was intro- duced around 1990 and reduced inci- dences by more than 95%, is based on a capsular polysaccharide isolated from the pathogen. The usage of synthetic antigens, and in particular carbohydrate conjugates in which the immunogen is synthesized chemically with atom scale precision, holds great promise for further developments along these lines (Astron- omo and Burton, 2010). This is predomi- nately due to the controlled production of a homogenous compound that minimizes batch-to-batch variability and thereby increases quality control standards. This might be accompanied with lower production costs in comparison with conven- tional vaccines. According to the World Health Organization (http://www.who.inV mediacentre/factsheets/fs294/en/index.

html), high costs are the cause for the slow introduction of the Hib vaccination in developing countries. A major leap forward in this field represents the devel- opment of the first synthetic conjugate polysaccharide vaccine that was reported 2004 by a Cuban-Canadian research team (Verez-Bencomo et aI., 2004). They first developed a high yielding route for the synthesis of a Hib polysaccharide

(4)

fragment, an oligomeric polyribosylribitol phosphate (Figure 3). These oligomers with an average of eight repeating units were repro- ducibly obtained in high yields (80%). The antigen was con- jugated to suitable carriers and immunogenicity studies performed in animals, adults,

"1ft:

o

OH t:H I

'--1>-1-1-

merases and, after incorporation of the acyclic nucleotide, further DNA synthesis is obvi- ated, a mechanism termed chain termination. In contrast to other antiviral nucleoside analogs, the phosphonates do not need the first phosphorylation step promoted by a kinase, which is often a bottleneck and counteracts the activity of the corn pounds.

children, and infants. After several clinical trials, a 99.7%

success rate was obtained, leading to commercial production as a vaccine (Quimi- Hib by Heber Biotech). This makes the polyribosylribitol

imatinib

}-NH

(~)--NH2

H°L---oj

acyc10vir

OH

Quimi-Hib

HPMPA _cidovofir

Chemistry & Biology invites your comments on this topic. Please write to the editors at chembiol@

cell.com.

phosphate conjugates to the first successful synthetic carbohydrate antigen. Given the great potential of medic- inal chemistry nowadays and,

Figure 3. Chemical Structure of Compounds Mentioned in the Main Text

REFERENCES

in particular, carbohydrate

chemistry, the approach of using synthetic antigens should allow the introduction of artificial carbohydrate analogs for increasing the immunogenicity of the conjugates in future.

(S)-1-(3-Hydroxy-2- phosphonylmethoxypropyl)

cytosine, Better Known as Cidofovir Compared to the structures that we have discussed before, cidofovir impresses by its structural simplicity (Figure 3). Cidofo- vir is an acyclic nucleoside phosphonate with very broad antiviral activity. Acyclic nucleosides may be described as nucleoside analogs with opened, simplified sugar surrogates. How was it discovered?

The history of acylic nucleoside analogs had its coming of age with the discovery of acyclovir as a selective inhibitor of herpes simplex virus (HSV) replication (De Clercq, 2008). In order to gain antiviral activity, nucleosides need to be phosphorylated to the triphosphates by three kinase-promoted steps. The resulting triphosphate is then used predominantly by the viral enzymes for incorporation of the modified nucleotide in the nascent DNA, and thus selectively impedes viral replication. Interestingly, although acyclovir is an acyclic guanosine analog, it was found by Elion et al. (1977) that the viral thymidine kinase more efficiently promotes the first phosphorylation step compared to the cellular enzymes. This makes acyclovir selective for virus-infected cells.

Further phosphorylation is promoted by the cellular kinases, resulting in the triphosphates that impede viral replication. In 1986, De Clercq, Holy, and colleagues reported on the discovery of a new class of broad-spectrum anti-DNA . virus agents: nucleoside phosphonates.

(S)-9-(3-Hydroxy-2-phosphonylmethoxypropyl)adenine (HPMPA) (Figure 3) is active against all tested DNA viruses (pol- yoma, papilloma, adeno, herpes, pox) (De Clercq et aI., 1986). Cidofovir is a closely related cytosine derivative of HPMPA that has been further developed clinically and has a remarkable broad-spectrum antiviral activity. It has been shown to be effi- cacious in the treatment of many DNA virus infections, including that of acyclovir- resistant HSV and varicella-zoster virus.

Interestingly, since cidofovir has shown promising effects in animal models, it was predicted to be effective in the prophylaxis and therapy of smallpox in hu- mans, a threat of bioterrorism. Thus, cidofovir is the sole antiviral drug that has been reserved and stockpiled for possible use in the therapy and prophylaxis of smallpox and the complications of smallpox vaccination (De Clercq and Holy, 2005).

Why are these phosphonates so effective? As depicted above, most nucleoside antiviral reagents get phosphorylated by host kinases in order to generate antiviral activity. The resulting triphosphates are selective substrates for the viral DNA poly-

Astronomo, R.D., and Burton, D.R.

(2010). Nat. Rev. Drug Discov. 9, 308--324.

Bierer, B.E., Somers, P.K., Wandless, T.J., Burak- off, S.J., and Schreiber, S.L. (1990). Science 250, 556-559.

Borisy, G.G., and Taylor, E.w. (1967a). J. Cell BioI.

34, 525-533.

Borisy, G.G., and Taylor, E.w. (1967b). J. Cell BioI.

34, 535-548.

Buchdunger, E., Zimmermann, J., Mett, H., Meyer, T., Muller, M., Druker, B.J., and Lydon, N.B. (1996).

Cancer Res. 56, 100-104.

Cohen, A., Cook, J., and Roe, E. (1940). Colchicine and related compounds (London: J. Chem. Soc.).

De Clereq, E. (2008). Med. Res. Rev. 28, 929-953.

De Clercq, E., and Holy, A. (2005). Nat. Rev. Drug Discov. 4, 928--940.

De Clercq, E., Holy, A., Rosenberg, I., Sakuma, T., Balzarini, J., and Maudgal, P.C. (1986). Nature 323, 464-467.

Desai, A., and Mitehison, T.J. (1997). Annu. Rev.

Cell Dev. BioI. 13,83-117.

Druker, B.J., Tamura, S., Buchdunger, E., Ohno, S., Segal, G.M., Fanning, S., Zimmermann, J., and Lydon, N.B. (1996). Nat. Med. 2,561-566.

Dustin, A. (1934). Bull. Aead. R. Med. Belg. 11, 187-502.

Eigsti, O.J. (1938). Proc. Natl. Aead. Sci. USA 24, 56-63.

Elion, G.B., Furman, PA, Fyfe, J.A., de Miranda, P., Beauehamp, L., and Sehaeffer, H.J. (1977).

Proc. Natl. Aead. Sei. USA 74, 5716--5720.

Engelborghs, Y. (1998). Eur. Biophys. J. 27, 437-445.

Goto, T., Kino, T., Hatanaka, H., Nishiyama, M., Okuhara, M., Kohsaka, M., Aoki, H., and Imanaka, H. (1987). Transplant. Proe. 19,4-8.

(5)

Harding, M.w., Galat, A., Uehling, D.E., and Schreiber, S.L. (1989). Nature 341,758-760.

Heisterkamp, N., Stam, K., Groffen, J., de Klein, A., and Grosveld, G. (1985). Nature 315,758-761.

Jordan, MA, and Wilson, L. (2004). Nat. Rev.

Cancer 4, 253-265.

Konopka, J.B., Watanabe, S.M., and Wilte, O.N.

(1984). Cell 37, 1035-1042.

Levine, M. (1945). Bot. Rev. 11, 145-180.

Liu, J., Farmer, J.D., Jr., Lane, W.S., Friedman, J., Weissman, I., and Schreiber, S.L. (1991). Cell 66, 807-815.

Mohri, H. (1968). Nature 217, 1053-1054.

Nowell, P.C. (1962). Blut8, 65-86.

Panda, D., Daijo, J.E., Jordan, MA, and Wilson, L.

(1995). Biochemistry 34,9921-9929.

Pelletier, and Caventou. (1820). Ann. Chim.

Physiol. 14, 69.

560

Pernice, B., and Caventou. (1889). Sicil. Med. 1, 265-279.

Powell, J.D., and Zheng, Y. (2006). Curr. Opin.

Investig. Drugs 7,1002-1007.

Ravelli, R.B., Gigant, B., Curmi, PA, Jourdain, I., Lachkar, S., Sobel, A., and Knossow, M. (2004).

Nature 428, 198-202.

Rosen, M.K., Standaert, R.F., Galat, A., Nakat- suka, M., and Schreiber, S.L. (1990). Science 248, 863-866.

Rowley, J.D. (1973). Nature 243,290-293.

Schindler, T., Bornmann, W., Pellicena, P., Miller, W.T., Clarkson, B., and Kuriyan, J. (2000). Science 289,1938-1942.

Shelanski, M.L., and Taylor, E.w. (1967). J. Cell BioI. 34, 549-554.

Shtivelman, E., Lifshitz, B., Gale, RP., and Canaani, E. (1985). Nature 315,550-554.

Siekierka, J.J., Hung, S.H., Poe, M., Lin, C.S., and Sigal, N.H. (1989). Nature 341,755-757.

Siekierka, J.J., Wiederrecht, G., Greulich, H., Boulton, D., Hung, S.H., Cryan, J., Hodges, P.J., and Sigal, N.H. (1990). J. Bioi. Chem. 265,

21011-21015. .

Standaert, RF., Galat, A., Verdine, G.L., and Schreiber, S.L. (1990). Nature 346,671-874.

Tanaka, H., Kuroda, A., Marusawa, H., Hashimoto, M., Hatanaka, H., Kino, T., Goto, T., and Okuhara, M. (1987). Transplant. Proc. 19,11-16.

Taylor, E.W. (1965). J. Cell BioI. 25 (Supp~, 145-160.

Tjio, J.H., and Levan, A. (1956). Hereditas 42, 1-8.

Verez-Bencomo, V., Fernandez-Santana, V., Hardy, E., Toledo, M.E., Rodriguez, M.C., Heynng- nezz, L., Rodriguez, A., Baly, A., Herrera, L., Izquierdo, M., et al. (2004). Science 305,522-525.

Wailer, C.F. (2010). Recent Results Cancer Res.

184,3-20.

Weisenberg, RC., Borisy, G.G., and Taylor, E.W.

(1968). Biochemistry 7, 4466-4479.

Five Molecules We Would Take to a Remote Island